feat: wildlife re-ID pipeline for offline species identification

[JasonWildMe]

Summary

Adds a complete offline wildlife re-identification pipeline to Off Grid Mobile, enabling field researchers to photograph animals, detect individuals via YOLO object detection, generate MiewID v4 embeddings, and match against pre-built species embedding packs — all without internet connectivity.

Key additions:

• Wildlife type system (species configs, detector configs, embedding packs, observations)
• ONNX Runtime inference service (stubs with TODO(P0) for real tensor ops)
• Detection preprocessing + YOLO postprocessing with NMS
• MiewID v4 embedding extraction + cosine similarity matching
• Pack manager for loading/indexing pre-built embedding databases
• Wildlife Zustand store with AsyncStorage persistence
• Full navigation structure: Home, Packs, Capture, Detection Results, Match Review, Observations, Sync
• Merged embedding database builder (pack + local individuals)
• App startup initialization for packs and wildlife store
• Integration tests for full pipeline flow and pack loading
• Stripped unused Off Grid modules (LLM, image gen, voice, tools, chat) to reduce bundle size

Type of Change

• New feature (non-breaking change that adds functionality)
• Refactor (code change that neither fixes a bug nor adds a feature)

Screenshots / Screen Recordings

N/A — no device testing yet; screens are structurally complete but require real ONNX inference wiring before visual validation.

Checklist

General

• My code follows the project's coding style and conventions
• I have performed a self-review of my code
• I have added/updated comments where the logic isn't self-evident
• My changes generate no new warnings or errors

Testing

• I have tested on Android (physical device or emulator)
• I have tested on iOS (physical device or simulator)
• I have tested in light mode and dark mode
• Existing tests pass locally (npm test) — 408 tests, 29 suites
• I have added tests that prove my fix is effective or my feature works

React Native Specific

• No new native module without corresponding platform implementation (Android + iOS)
• No hardcoded pixel values — uses SPACING / TYPOGRAPHY constants from the theme
• Styles use useThemedStyles pattern (not inline or static StyleSheet.create)
• Large lists use FlatList / FlashList (not .map() inside ScrollView)

Performance & Models

• File paths are resolved correctly on both platforms (no hardcoded / vs \\)
• Large files (models, assets) are not committed to the repository

Security

• No secrets, API keys, or credentials are included in the code
• User input is validated/sanitized where applicable

Related Issues

This implements the wildlife re-ID feasibility plan documented in docs/plans/2026-02-25-wildlife-reid-implementation.md.

Additional Notes

• ONNX inference methods are currently stubs (marked TODO(P0)) — they return empty tensors. Wiring real image-to-tensor conversion and ONNX session execution is the next step.
• The sync screen shows a queue UI but sync functionality is not yet implemented (shows an alert).
• Theme palette now includes statusSuccess, statusWarning, statusError for dark-mode-aware status indicators.
• DetectorConfig loading falls back to safe YOLO defaults when pack config isn't available.